tensornn.loss.MeanSquaredError

class tensornn.loss.MeanSquaredError

Bases: Loss

Mean squared error is calculated extremely simply. 1. Find the difference between the prediction vs. the actual results we should have got 2. Square these values, because negatives are the same as positives, only magnitude matters 3. calculate mean

ex: our predictions: [0.1, 0.2, 0.7], desired: [0, 0, 1] 1. pred - actual: [0.1, 0.2, -0.3] 2. squared: [0.01, 0.04, 0.09] 3. mean: 0.04666667

Methods

calculate	The loss function is used to calculate how off the predictions of the network are.
derivative	Used in backpropagation which helps calculates how much each neuron impacts the loss.
source

__repr__()

Return repr(self).

calculate(pred: Tensor, desired: Tensor) → Tensor

The loss function is used to calculate how off the predictions of the network are.

Parameters:

• pred – the prediction of the network

• desired – the desired values which the network should have gotten close to

Returns:

the average of calculated loss for one whole pass of the network

derivative(pred: Tensor, desired: Tensor) → Tensor

Used in backpropagation which helps calculates how much each neuron impacts the loss.

Parameters:

• pred – the prediction of the network

• desired – the desired values which the network should have gotten close to

Returns:

the derivative of the loss function wrt the last layer of the network